The discovery that Green Fluorescent Protein (GFP) from the jellyfish A. victoria retains its fluorescent properties when expressed in heterologous cells has provided biological research with a new, unique and powerful tool (Chalfie et al (1994). Science 263:802; Prasher (1995) Trends in Genetics 11:320; WO 95/07463).
Furthermore, the discovery of a blue fluorescent variant of GFP (Heim et al. (1994). Proc.Natl.Acad.Sci. 91:12501) has greatly increased the potential applications of using fluorescent recombinant probes to monitor cellular events or functions, since the availability of probes having different excitation and emission spectra permits simultaneous monitoring of more than one process.
However, the blue fluorescing variant described by Heim et al, Y66H-GFP, suffers from certain limitations: The blue fluorescence is weak (emission maximum at 448 nm), thus making detection difficult, and necessitating prolonged excitation of cells expressing Y66H-GFP. Moreover, the prolonged period of excitation is damaging to cells especially because the excitation wavelength is in the UV range, 360 nm-390 nm.
A very important aspect of using recombinant, fluorescent proteins in studying cellular functions is the non-invasive nature of the assay. This allows detection of cellular events in intact, living cells. A limitation with current fluorescent proteins is, however, that relatively high intensity light sources are needed for visualization. Especially with the blue variant, Y66H-GFP, it is necessary to excite with intensities that are damaging to most cells. It is worth mentioning that some cellular events like oscillations in intracellular signalling systems, e.g. cytosolic free calcium, are very photo sensitive. A further consequence of the low light emittance is that only high levels of expression can be detected. Obtaining such high level expression may stress the transcriptional and/or translational machinery of the cells.
The excitation spectrum of the green fluorescent protein from Aequorea Victoria shows two peaks: A major peak at 396 nm, which is in the potentially cell damaging UV range, and a lesser peak at 475 nm, which is in an excitation range that is much less harmful to cells. Heim et al.(1995), Nature, Vol. 373, p. 663-4, discloses a Ser65Thr mutation of GFP (S65T) having longer wavelengths of excitation and emission, 490 nm and 510 nm, respectively, than the wild-type GFP and wherein the fluorophore formation proceeded about fourfold more rapidly than in the wild-type GFP.
Expression of GFP or its fluorescent variants in living cells provides a valuable tool for studying cellular events and it is well known that many cells, including mammalian cells, are incubated at approximately 37.degree. C. in order to secure optimal and/or physiologically relevant growth. Cell lines originating from different organisms or tissues may have different relevant temperatures ranging from about 35.degree. C. for fibroblasts to about 38.degree. C.-39.degree. C. for mouse .beta.-cells. Experience has shown, however, that the fluorescent signal from cells expressing GFP is weak or absent when said cells are incubated at temperatures above room temperature, cf. Webb, C. D. et al., Journal of Bacteriology, October 1995, p. 5906-5911. Ogawa H. et al., Proc. Natl. Acad. Sci. USA, Vol. 92, pp. 11899-11903, December 1995, and Lim et al. J. Biochem. 118, 13-17 (1995). The improved fluorescent variant S65T described by Heim et al. (1995) supra also displays very low fluorescence when incubated under normal culture conditions (37.degree. C.), cf. Kaether and Gerdes FEBS Letters 369 (1995) pp. 267-271. Many experiments involving the study of cell metabolism are dependent on the possibility of incubating the cells at physiologically relevant temperatures, i.e. temperatures at about 37.degree. C.